X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Futils.py;h=b644f4e920bf0353658ec9920abdb0541dbaf0e2;hb=0003a5c4163df6b5a7fd90ec256ea7497f639dda;hp=d4df78071663489752a80a1a833662fe5505ec3b;hpb=478c2c619364f5fb0c1ee9e9489048ab4ae26521;p=youtube-dl diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py index d4df78071..b644f4e92 100644 --- a/youtube_dl/utils.py +++ b/youtube_dl/utils.py @@ -24,6 +24,7 @@ import socket import struct import subprocess import sys +import tempfile import traceback import xml.etree.ElementTree import zlib @@ -91,11 +92,9 @@ except ImportError: compat_subprocess_get_DEVNULL = lambda: open(os.path.devnull, 'w') try: - from urllib.parse import parse_qs as compat_parse_qs -except ImportError: # Python 2 - # HACK: The following is the correct parse_qs implementation from cpython 3's stdlib. - # Python 2's version is apparently totally broken - def _unquote(string, encoding='utf-8', errors='replace'): + from urllib.parse import unquote as compat_urllib_parse_unquote +except ImportError: + def compat_urllib_parse_unquote(string, encoding='utf-8', errors='replace'): if string == '': return string res = string.split('%') @@ -130,6 +129,13 @@ except ImportError: # Python 2 string += pct_sequence.decode(encoding, errors) return string + +try: + from urllib.parse import parse_qs as compat_parse_qs +except ImportError: # Python 2 + # HACK: The following is the correct parse_qs implementation from cpython 3's stdlib. + # Python 2's version is apparently totally broken + def _parse_qsl(qs, keep_blank_values=False, strict_parsing=False, encoding='utf-8', errors='replace'): qs, _coerce_result = qs, unicode @@ -149,10 +155,12 @@ except ImportError: # Python 2 continue if len(nv[1]) or keep_blank_values: name = nv[0].replace('+', ' ') - name = _unquote(name, encoding=encoding, errors=errors) + name = compat_urllib_parse_unquote( + name, encoding=encoding, errors=errors) name = _coerce_result(name) value = nv[1].replace('+', ' ') - value = _unquote(value, encoding=encoding, errors=errors) + value = compat_urllib_parse_unquote( + value, encoding=encoding, errors=errors) value = _coerce_result(value) r.append((name, value)) return r @@ -184,6 +192,13 @@ try: except ImportError: # Python 2.6 from xml.parsers.expat import ExpatError as compat_xml_parse_error +try: + from shlex import quote as shlex_quote +except ImportError: # Python < 3.3 + def shlex_quote(s): + return "'" + s.replace("'", "'\"'\"'") + "'" + + def compat_ord(c): if type(c) is int: return c else: return ord(c) @@ -221,26 +236,55 @@ else: assert type(s) == type(u'') print(s) -# In Python 2.x, json.dump expects a bytestream. -# In Python 3.x, it writes to a character stream -if sys.version_info < (3,0): - def write_json_file(obj, fn): - with open(fn, 'wb') as f: - json.dump(obj, f) -else: - def write_json_file(obj, fn): - with open(fn, 'w', encoding='utf-8') as f: - json.dump(obj, f) -if sys.version_info >= (2,7): +def write_json_file(obj, fn): + """ Encode obj as JSON and write it to fn, atomically """ + + args = { + 'suffix': '.tmp', + 'prefix': os.path.basename(fn) + '.', + 'dir': os.path.dirname(fn), + 'delete': False, + } + + # In Python 2.x, json.dump expects a bytestream. + # In Python 3.x, it writes to a character stream + if sys.version_info < (3, 0): + args['mode'] = 'wb' + else: + args.update({ + 'mode': 'w', + 'encoding': 'utf-8', + }) + + tf = tempfile.NamedTemporaryFile(**args) + + try: + with tf: + json.dump(obj, tf) + os.rename(tf.name, fn) + except: + try: + os.remove(tf.name) + except OSError: + pass + raise + + +if sys.version_info >= (2, 7): def find_xpath_attr(node, xpath, key, val): """ Find the xpath xpath[@key=val] """ - assert re.match(r'^[a-zA-Z]+$', key) - assert re.match(r'^[a-zA-Z0-9@\s:._]*$', val) + assert re.match(r'^[a-zA-Z-]+$', key) + assert re.match(r'^[a-zA-Z0-9@\s:._-]*$', val) expr = xpath + u"[@%s='%s']" % (key, val) return node.find(expr) else: def find_xpath_attr(node, xpath, key, val): + # Here comes the crazy part: In 2.6, if the xpath is a unicode, + # .//node does not match if a node is a direct child of . ! + if isinstance(xpath, unicode): + xpath = xpath.encode('ascii') + for f in node.findall(xpath): if f.attrib.get(key) == val: return f @@ -259,30 +303,20 @@ def xpath_with_ns(path, ns_map): replaced.append('{%s}%s' % (ns_map[ns], tag)) return '/'.join(replaced) -def htmlentity_transform(matchobj): - """Transforms an HTML entity to a character. - This function receives a match object and is intended to be used with - the re.sub() function. - """ - entity = matchobj.group(1) +def xpath_text(node, xpath, name=None, fatal=False): + if sys.version_info < (2, 7): # Crazy 2.6 + xpath = xpath.encode('ascii') - # Known non-numeric HTML entity - if entity in compat_html_entities.name2codepoint: - return compat_chr(compat_html_entities.name2codepoint[entity]) - - mobj = re.match(u'(?u)#(x?\\d+)', entity) - if mobj is not None: - numstr = mobj.group(1) - if numstr.startswith(u'x'): - base = 16 - numstr = u'0%s' % numstr + n = node.find(xpath) + if n is None: + if fatal: + name = xpath if name is None else name + raise ExtractorError('Could not find XML element %s' % name) else: - base = 10 - return compat_chr(int(numstr, base)) + return None + return n.text - # Unknown entity in name, return its literal representation - return (u'&%s;' % entity) compat_html_parser.locatestarttagend = re.compile(r"""<[a-zA-Z][-.a-zA-Z0-9:_]*(?:\s+(?:(?<=['"\s])[^\s/>][^\s/=>]*(?:\s*=+\s*(?:'[^']*'|"[^"]*"|(?!['"])[^>\s]*))?\s*)*)?\s*""", re.VERBOSE) # backport bugfix class BaseHTMLParser(compat_html_parser.HTMLParser): @@ -504,13 +538,33 @@ def orderedSet(iterable): return res +def _htmlentity_transform(entity): + """Transforms an HTML entity to a character.""" + # Known non-numeric HTML entity + if entity in compat_html_entities.name2codepoint: + return compat_chr(compat_html_entities.name2codepoint[entity]) + + mobj = re.match(r'#(x?[0-9]+)', entity) + if mobj is not None: + numstr = mobj.group(1) + if numstr.startswith(u'x'): + base = 16 + numstr = u'0%s' % numstr + else: + base = 10 + return compat_chr(int(numstr, base)) + + # Unknown entity in name, return its literal representation + return (u'&%s;' % entity) + + def unescapeHTML(s): if s is None: return None assert type(s) == compat_str - result = re.sub(r'(?u)&(.+?);', htmlentity_transform, s) - return result + return re.sub( + r'&([^;]+);', lambda m: _htmlentity_transform(m.group(1)), s) def encodeFilename(s, for_subprocess=False): @@ -540,6 +594,16 @@ def encodeFilename(s, for_subprocess=False): encoding = 'utf-8' return s.encode(encoding, 'ignore') + +def encodeArgument(s): + if not isinstance(s, compat_str): + # Legacy code that uses byte strings + # Uncomment the following line after fixing all post processors + #assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s)) + s = s.decode('ascii') + return encodeFilename(s, True) + + def decodeOption(optval): if optval is None: return optval @@ -572,7 +636,7 @@ def make_HTTPS_handler(opts_no_check_certificate, **kwargs): self.sock = sock self._tunnel() try: - self.sock = ssl.wrap_socket(sock, self.key_file, self.cert_file, ssl_version=ssl.PROTOCOL_SSLv3) + self.sock = ssl.wrap_socket(sock, self.key_file, self.cert_file, ssl_version=ssl.PROTOCOL_TLSv1) except ssl.SSLError: self.sock = ssl.wrap_socket(sock, self.key_file, self.cert_file, ssl_version=ssl.PROTOCOL_SSLv23) @@ -580,8 +644,14 @@ def make_HTTPS_handler(opts_no_check_certificate, **kwargs): def https_open(self, req): return self.do_open(HTTPSConnectionV3, req) return HTTPSHandlerV3(**kwargs) - else: - context = ssl.SSLContext(ssl.PROTOCOL_SSLv3) + elif hasattr(ssl, 'create_default_context'): # Python >= 3.4 + context = ssl.create_default_context(ssl.Purpose.CLIENT_AUTH) + context.options &= ~ssl.OP_NO_SSLv3 # Allow older, not-as-secure SSLv3 + if opts_no_check_certificate: + context.verify_mode = ssl.CERT_NONE + return compat_urllib_request.HTTPSHandler(context=context, **kwargs) + else: # Python < 3.4 + context = ssl.SSLContext(ssl.PROTOCOL_SSLv23) context.verify_mode = (ssl.CERT_NONE if opts_no_check_certificate else ssl.CERT_REQUIRED) @@ -594,13 +664,15 @@ def make_HTTPS_handler(opts_no_check_certificate, **kwargs): class ExtractorError(Exception): """Error during info extraction.""" - def __init__(self, msg, tb=None, expected=False, cause=None): + def __init__(self, msg, tb=None, expected=False, cause=None, video_id=None): """ tb, if given, is the original traceback (so that it can be printed out). If expected is set, this is a normal error message and most likely not a bug in youtube-dl. """ if sys.exc_info()[0] in (compat_urllib_error.URLError, socket.timeout, UnavailableVideoError): expected = True + if video_id is not None: + msg = video_id + ': ' + msg if not expected: msg = msg + u'; please report this issue on https://yt-dl.org/bug . Be sure to call youtube-dl with the --verbose flag and include its complete output. Make sure you are using the latest version; type youtube-dl -U to update.' super(ExtractorError, self).__init__(msg) @@ -608,6 +680,7 @@ class ExtractorError(Exception): self.traceback = tb self.exc_info = sys.exc_info() # preserve original exception self.cause = cause + self.video_id = video_id def format_traceback(self): if self.traceback is None: @@ -714,10 +787,9 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler): return ret def http_request(self, req): - for h,v in std_headers.items(): - if h in req.headers: - del req.headers[h] - req.add_header(h, v) + for h, v in std_headers.items(): + if h not in req.headers: + req.add_header(h, v) if 'Youtubedl-no-compression' in req.headers: if 'Accept-encoding' in req.headers: del req.headers['Accept-encoding'] @@ -762,7 +834,7 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler): https_response = http_response -def parse_iso8601(date_str): +def parse_iso8601(date_str, delimiter='T'): """ Return a UNIX timestamp from the given date """ if date_str is None: @@ -782,8 +854,8 @@ def parse_iso8601(date_str): timezone = datetime.timedelta( hours=sign * int(m.group('hours')), minutes=sign * int(m.group('minutes'))) - - dt = datetime.datetime.strptime(date_str, '%Y-%m-%dT%H:%M:%S') - timezone + date_format = '%Y-%m-%d{0}%H:%M:%S'.format(delimiter) + dt = datetime.datetime.strptime(date_str, date_format) - timezone return calendar.timegm(dt.timetuple()) @@ -803,9 +875,14 @@ def unified_strdate(date_str): '%d %b %Y', '%B %d %Y', '%b %d %Y', + '%b %dst %Y %I:%M%p', + '%b %dnd %Y %I:%M%p', + '%b %dth %Y %I:%M%p', '%Y-%m-%d', + '%Y/%m/%d', '%d.%m.%Y', '%d/%m/%Y', + '%d/%m/%y', '%Y/%m/%d %H:%M:%S', '%Y-%m-%d %H:%M:%S', '%d.%m.%Y %H:%M', @@ -829,6 +906,8 @@ def unified_strdate(date_str): return upload_date def determine_ext(url, default_ext=u'unknown_video'): + if url is None: + return default_ext guess = url.partition(u'?')[0].rpartition(u'.')[2] if re.match(r'^[A-Za-z0-9]+$', guess): return guess @@ -923,7 +1002,11 @@ def _windows_write_string(s, out): 2: -12, } - fileno = out.fileno() + try: + fileno = out.fileno() + except AttributeError: + # If the output stream doesn't have a fileno, it's virtual + return False if fileno not in WIN_OUTPUT_IDS: return False @@ -1018,12 +1101,6 @@ def intlist_to_bytes(xs): return bytes(xs) -def get_cachedir(params={}): - cache_root = os.environ.get('XDG_CACHE_HOME', - os.path.expanduser('~/.cache')) - return params.get('cachedir', os.path.join(cache_root, 'youtube-dl')) - - # Cross-platform file locking if sys.platform == 'win32': import ctypes.wintypes @@ -1083,10 +1160,10 @@ else: import fcntl def _lock_file(f, exclusive): - fcntl.lockf(f, fcntl.LOCK_EX if exclusive else fcntl.LOCK_SH) + fcntl.flock(f, fcntl.LOCK_EX if exclusive else fcntl.LOCK_SH) def _unlock_file(f): - fcntl.lockf(f, fcntl.LOCK_UN) + fcntl.flock(f, fcntl.LOCK_UN) class locked_file(object): @@ -1173,11 +1250,6 @@ def format_bytes(bytes): return u'%.2f%s' % (converted, suffix) -def str_to_int(int_str): - int_str = re.sub(r'[,\.]', u'', int_str) - return int(int_str) - - def get_term_width(): columns = os.environ.get('COLUMNS', None) if columns: @@ -1235,6 +1307,12 @@ def remove_start(s, start): return s +def remove_end(s, end): + if s.endswith(end): + return s[:-len(end)] + return s + + def url_basename(url): path = compat_urlparse.urlparse(url).path return path.strip(u'/').split(u'/')[-1] @@ -1245,20 +1323,39 @@ class HEADRequest(compat_urllib_request.Request): return "HEAD" -def int_or_none(v, scale=1, default=None): - return default if v is None else (int(v) // scale) +def int_or_none(v, scale=1, default=None, get_attr=None, invscale=1): + if get_attr: + if v is not None: + v = getattr(v, get_attr, None) + if v == '': + v = None + return default if v is None else (int(v) * invscale // scale) -def float_or_none(v, scale=1, default=None): - return default if v is None else (float(v) / scale) +def str_or_none(v, default=None): + return default if v is None else compat_str(v) + + +def str_to_int(int_str): + """ A more relaxed version of int_or_none """ + if int_str is None: + return None + int_str = re.sub(r'[,\.\+]', u'', int_str) + return int(int_str) + + +def float_or_none(v, scale=1, invscale=1, default=None): + return default if v is None else (float(v) * invscale / scale) def parse_duration(s): if s is None: return None + s = s.strip() + m = re.match( - r'(?:(?:(?P[0-9]+)[:h])?(?P[0-9]+)[:m])?(?P[0-9]+)s?(?::[0-9]+)?$', s) + r'(?i)(?:(?:(?P[0-9]+)\s*(?:[:h]|hours?)\s*)?(?P[0-9]+)\s*(?:[:m]|mins?|minutes?)\s*)?(?P[0-9]+)(?P\.[0-9]+)?\s*(?:s|secs?|seconds?)?$', s) if not m: return None res = int(m.group('secs')) @@ -1266,6 +1363,8 @@ def parse_duration(s): res += int(m.group('mins')) * 60 if m.group('hours'): res += int(m.group('hours')) * 60 * 60 + if m.group('ms'): + res += float(m.group('ms')) return res @@ -1338,6 +1437,24 @@ def uppercase_escape(s): lambda m: unicode_escape(m.group(0))[0], s) + +def escape_rfc3986(s): + """Escape non-ASCII characters as suggested by RFC 3986""" + if sys.version_info < (3, 0) and isinstance(s, unicode): + s = s.encode('utf-8') + return compat_urllib_parse.quote(s, "%/;:@&=+$,!~*'()?#[]") + + +def escape_url(url): + """Escape URL as suggested by RFC 3986""" + url_parsed = compat_urllib_parse_urlparse(url) + return url_parsed._replace( + path=escape_rfc3986(url_parsed.path), + params=escape_rfc3986(url_parsed.params), + query=escape_rfc3986(url_parsed.query), + fragment=escape_rfc3986(url_parsed.fragment) + ).geturl() + try: struct.pack(u'!I', 0) except TypeError: @@ -1376,6 +1493,12 @@ def urlencode_postdata(*args, **kargs): return compat_urllib_parse.urlencode(*args, **kargs).encode('ascii') +try: + etree_iter = xml.etree.ElementTree.Element.iter +except AttributeError: # Python <=2.6 + etree_iter = lambda n: n.findall('.//*') + + def parse_xml(s): class TreeBuilder(xml.etree.ElementTree.TreeBuilder): def doctype(self, name, pubid, system): @@ -1383,7 +1506,14 @@ def parse_xml(s): parser = xml.etree.ElementTree.XMLParser(target=TreeBuilder()) kwargs = {'parser': parser} if sys.version_info >= (2, 7) else {} - return xml.etree.ElementTree.XML(s.encode('utf-8'), **kwargs) + tree = xml.etree.ElementTree.XML(s.encode('utf-8'), **kwargs) + # Fix up XML parser in Python 2.x + if sys.version_info < (3, 0): + for n in etree_iter(tree): + if n.text is not None: + if not isinstance(n.text, compat_str): + n.text = n.text.decode('utf-8') + return tree if sys.version_info < (3, 0) and sys.platform == 'win32': @@ -1405,7 +1535,35 @@ US_RATINGS = { def strip_jsonp(code): - return re.sub(r'(?s)^[a-zA-Z_]+\s*\(\s*(.*)\);\s*?\s*$', r'\1', code) + return re.sub(r'(?s)^[a-zA-Z0-9_]+\s*\(\s*(.*)\);?\s*?\s*$', r'\1', code) + + +def js_to_json(code): + def fix_kv(m): + key = m.group(2) + if key.startswith("'"): + assert key.endswith("'") + assert '"' not in key + key = '"%s"' % key[1:-1] + elif not key.startswith('"'): + key = '"%s"' % key + + value = m.group(4) + if value.startswith("'"): + assert value.endswith("'") + assert '"' not in value + value = '"%s"' % value[1:-1] + + return m.group(1) + key + m.group(3) + value + + res = re.sub(r'''(?x) + ([{,]\s*) + ("[^"]*"|\'[^\']*\'|[a-z0-9A-Z]+) + (:\s*) + ([0-9.]+|true|false|"[^"]*"|\'[^\']*\'|\[|\{) + ''', fix_kv, code) + res = re.sub(r',(\s*\])', lambda m: m.group(1), res) + return res def qualities(quality_ids): @@ -1417,3 +1575,27 @@ def qualities(quality_ids): return -1 return q + +DEFAULT_OUTTMPL = '%(title)s-%(id)s.%(ext)s' + +try: + subprocess_check_output = subprocess.check_output +except AttributeError: + def subprocess_check_output(*args, **kwargs): + assert 'input' not in kwargs + p = subprocess.Popen(*args, stdout=subprocess.PIPE, **kwargs) + output, _ = p.communicate() + ret = p.poll() + if ret: + raise subprocess.CalledProcessError(ret, p.args, output=output) + return output + + +def limit_length(s, length): + """ Add ellipses to overly long strings """ + if s is None: + return None + ELLIPSES = '...' + if len(s) > length: + return s[:length - len(ELLIPSES)] + ELLIPSES + return s